Abstract
An item analysis of the MMPIs of suicidal (commits, attempts, and threats) and non-suicidal (controls) neuropsychiatric hospital patients was carried out and then replicated. Significant differences appeared consistently only between the threat-suicide group and all other groups. An experimental form of a “threat suicide scale” made up of 52 items which differentiated between threat-suicides and non-suicides in both an original and a replication study was constructed from 15 items significant at the .01 level or better, an additional 24 items at the .05 level, and another 13 items at the .10 level. Application of the scale to a prediction sample indicated an increase in predictive efficiency over use of the base rate alone. The scale also provides possibilities for increased understanding of suicidal activity.
Get full access to this article
View all access options for this article.